Beschreibung:
<jats:p><jats:italic>Context</jats:italic>. One of the most important properties of a galaxy is the total stellar mass, or equivalently the stellar mass-to-light ratio (<jats:italic>M</jats:italic>/<jats:italic>L</jats:italic>). It is not directly observable, but can be estimated from stellar population synthesis. Currently, a galaxy’s <jats:italic>M</jats:italic>/<jats:italic>L</jats:italic> is typically estimated from global fluxes. For example, a single global <jats:italic>g</jats:italic> − <jats:italic>i</jats:italic> colour correlates well with the stellar <jats:italic>M</jats:italic>/<jats:italic>L</jats:italic>. Spectral energy distribution (SED) fitting can make use of all available fluxes and their errors to make a Bayesian estimate of the <jats:italic>M</jats:italic>/<jats:italic>L</jats:italic>.</jats:p>
<jats:p><jats:italic>Aims</jats:italic>. We want to investigate the possibility of using morphology information to assist predictions of <jats:italic>M</jats:italic>/<jats:italic>L</jats:italic>. Our first goal is to develop and train a method that only requires a <jats:italic>g</jats:italic>-band image and redshift as input. This will allows us to study the correlation between <jats:italic>M</jats:italic>/<jats:italic>L</jats:italic> and morphology. Next, we can also include the <jats:italic>i</jats:italic>-band flux, and determine if morphology provides additional constraints compared to a method that only uses <jats:italic>g</jats:italic>- and <jats:italic>i</jats:italic>-band fluxes.</jats:p>
<jats:p><jats:italic>Methods</jats:italic>. We used a machine learning pipeline that can be split in two steps. First, we detected morphology features with a convolutional neural network. These are then combined with redshift, pixel size and <jats:italic>g</jats:italic>-band luminosity features in a gradient boosting machine. Our training target was the <jats:italic>M</jats:italic>/<jats:italic>L</jats:italic> acquired from the GALEX-SDSS-WISE Legacy Catalog, which uses global SED fitting and contains galaxies with <jats:italic>z</jats:italic> ∼ 0.1.</jats:p>
<jats:p><jats:italic>Results</jats:italic>. Morphology is a useful attribute when no colour information is available, but can not outperform colour methods on its own. When we combine the morphology features with global <jats:italic>g</jats:italic>- and <jats:italic>i</jats:italic>-band luminosities, we find an improved estimate compared to a model which does not make use of morphology.</jats:p>
<jats:p><jats:italic>Conclusions</jats:italic>. While our method was trained to reproduce global SED fitted <jats:italic>M</jats:italic>/<jats:italic>L</jats:italic>, galaxy morphology gives us an important additional constraint when using one or two bands. Our framework can be extended to other problems to make use of morphological information.</jats:p>